When the user searches the Internet, CD-ROM, or the like for a target product, and wants to visually confirm it, the image, color, and the like of the product can be clearly displayed on a color display. However, the uneven shape of the product surface can be expressed to some extent on an image by, e.g., shading upon taking its picture, but the uneven shape that can be identified on the color display is far from satisfactory. For this reason, it is nearly impossible to accurately examine material impression such as an uneven shape and the like.
As a method of three-dimensionally reconstructing the uneven shape of an object, and outputting the reconstructed shape to be identifiable by the user, a technique that transmits tactile information to the user by pressure actuators which are arranged in an array on a silicon substrate (Japanese Patent Laid-Open No. 5-216568), and a technique that transmits tactile information such as unevenness information to the user by three-dimensionally deforming or vibrating a piezoelectric sheet (Japanese Patent Laid-Open No. 2000-148393) are known.
However, the invention disclosed in Japanese Patent Laid-Open No. 5-216568 arranges the actuators on the silicon substrate to express tactile information but does not consider image display at all. Hence, tactile information such as unevenness information or the like cannot be output to be superposed on image display. For this reason, it is impossible for this method to display an image and its unevenness information as combined information.
Japanese Patent Laid-Open No. 2000-148393 describes an invention relating to an output apparatus which presents tactile information such as unevenness, vibrations, and the like to be superposed on image display using a transparent piezoelectric material. However, with this method, tactile information is obtained as follows. That is, the second derivative of luminance of a displayed image is calculated, contour information of the image is acquired by extracting the second derivative of luminance, and the inner and outer portions of the contour are merely output as differences of unevenness. Hence, this method cannot finely output the texture shape of the displayed object surface. Therefore, it is impossible for both the conventional methods to output texture impression based on the uneven shape of an object together with an image of that object.